Operational definitions of two key variables in group communication can provide a theoretical perspective for considering communication behavior in group research: "risk taking," the tendency to prefer long shots with higher payoffs over sure things with lower payoffs; and the "risky-shift" phenomenon, the finding that after group discussion the individual members will recommend a less conservative course of action than they had espoused before. This paper reviews some of the literature relevant to these variables, and recommends guidelines for the use of the "risky-shift" phenomenon as the theoretic rationale for future research. (Author/SH)